koros0111/Warrior_EA
1
0
Fork 0
forked from animatedread/Warrior_EA
Code Pull requests Activity

Commit graph

  • e8452913c0 feat: add swing-context feature with confirmed zigzag pivot main AnimateDread 2026-07-19 11:04:38 -04:00
  • 3fbe16127e docs: simplify comments and enum descriptions across AI and Enums AnimateDread 2026-07-18 23:59:40 -04:00
  • 6f95a402f4 refactor: replace class-balance loss weighting with data-level oversampling AnimateDread 2026-07-18 23:25:54 -04:00
  • 39d785b0de fix: prevent oversampling weight compounding and fix era display lag AnimateDread 2026-07-18 23:03:51 -04:00
  • 20f32306f3 feat(ml): add data-level oversampling for minority classes in training queue AnimateDread 2026-07-18 22:27:34 -04:00
  • 8ce30f39c3 . AnimateDread 2026-07-18 17:43:29 -04:00
  • d91cabc114 refactor(MoneyIntelligent): replace streak-chasing lot sizing with fractional-Kelly criterion AnimateDread 2026-07-18 17:39:58 -04:00
  • 31b2711c8f feat: add daily-loss and max-drawdown risk circuit breakers AnimateDread 2026-07-18 17:29:38 -04:00
  • d75bd6d6a8 feat: add configurable news event proximity/impact as an NN input feature AnimateDread 2026-07-18 17:26:04 -04:00
  • d0e89a6fc7 fix(SignalNewsFilter): scope calendar veto to the traded symbol's own currencies AnimateDread 2026-07-18 17:22:33 -04:00
  • 4c0afd682b refactor(ExpertSignalAIBase): extract AutoTune param state into CADIndicatorTuner AnimateDread 2026-07-18 16:22:09 -04:00
  • e4a154aa5e refactor(AI): split 8 self-contained classes out of the Network.mqh god-file AnimateDread 2026-07-18 16:13:03 -04:00
  • e6ff2506a8 refactor(Money): extract CMoneyRiskBase to remove FixedRisk/Intelligent duplication AnimateDread 2026-07-18 15:53:04 -04:00
  • 97cf7be556 chore: cleanup pass - dead code, stale wizard scaffolding, one edge-case guard AnimateDread 2026-07-18 15:49:59 -04:00
  • a9b497f615 fix(Inputs): shorten input comments that overflowed MQL5's 255-char identifier=value limit AnimateDread 2026-07-18 15:44:57 -04:00
  • 7e27ad8729 fix(ExpertSignalAIBase): stop pass-3 resume from replaying pass 2 forever AnimateDread 2026-07-18 15:44:48 -04:00
  • 6a687cda41 feat: add SGD+momentum optimizer and input-driven hyperparameters AnimateDread 2026-07-18 14:56:41 -04:00
  • 86cb37595d feat: introduce Pass 3 for OOS scoring and extract status label AnimateDread 2026-07-18 12:10:37 -04:00
  • 379db0827c fix(ExpertSignalAIBase): add shuffled backprop pass 2 to prevent training instability AnimateDread 2026-07-18 11:33:21 -04:00
  • 87ca08f21b refactor: tune optimizer momentum and integrate focal loss to reduce multi-era bias AnimateDread 2026-07-18 10:03:21 -04:00
  • 1a804fce28 feat: make class-balance oversampling multiplier configurable via input enum AnimateDread 2026-07-18 02:01:06 -04:00
  • 457ae4acac refactor: Remove test holdout slice, add live signal alternation gate AnimateDread 2026-07-17 23:55:10 -04:00
  • 5e97bded86 chore: update DirectML binaries (CPU and DML) with new compiled output AnimateDread 2026-07-17 23:23:15 -04:00
  • e62c710d6f fix: correct array orientation and PReLU gradient backprop in hidden layers AnimateDread 2026-07-17 23:21:12 -04:00
  • e50f7d6c3b refactor: improve status label formatting in training schedule AnimateDread 2026-07-17 22:05:52 -04:00
  • 58049461c0 feat: show in-sample training percentage in progress log AnimateDread 2026-07-17 21:56:25 -04:00
  • ef95e19d1d fix: correct metric label in training output and improve status panel comments AnimateDread 2026-07-17 21:53:09 -04:00
  • fe4c93cf71 refactor: simplify on-chart status text and remove verbose metrics AnimateDread 2026-07-17 21:49:30 -04:00
  • 6f3daf23a3 fix(System/StatusLabel): correct type mismatches in TextGetSize usage AnimateDread 2026-07-17 21:37:52 -04:00
  • a2ed4e3682 fix: word-wrap status labels via TextGetSize() to fix clipping and background overflow AnimateDread 2026-07-17 21:36:44 -04:00
  • bb9edfc119 fix(status-label): handle multi-line status text with per-line labels AnimateDread 2026-07-17 21:32:55 -04:00
  • 2c0cdf5e2b refactor(ai): replace Comment() with StatusLabel for training status display AnimateDread 2026-07-17 21:28:59 -04:00
  • f5fa5b3fcc fix: always prebuild label cache to prevent silent neutral labels on checkpoint restart AnimateDread 2026-07-17 21:18:25 -04:00
  • 55ff1c40c4 perf: improve small layer dispatch and UI responsiveness AnimateDread 2026-07-17 19:30:10 -04:00
  • 5e9e2f4136 perf: switch OpenCL kernels to float32 for consumer GPU compatibility AnimateDread 2026-07-17 19:04:56 -04:00
  • b034430ed5 feat: switch weight initialization from LeCun-uniform to He-uniform AnimateDread 2026-07-17 14:51:47 -04:00
  • 5a25771d20 fix: handle additional OCL neuron types and increase indicator period for swing detection AnimateDread 2026-07-17 14:34:00 -04:00
  • 3faba25fbe AnimateDread 2026-07-17 09:29:45 -04:00
  • 80a53d5b04 feat: cap class sample weight and increase training time budget AnimateDread 2026-07-17 09:11:42 -04:00
  • ed2285000b feat: widen label cache to include neighborhood around pivot bars AnimateDread 2026-07-16 22:29:49 -04:00
  • 17abf513ca feat: bias output layer toward the real class prior to fix era-0 cold-start noise AnimateDread 2026-07-16 20:37:36 -04:00
  • 721f2970f2 feat: normalize CumulativeDelta feature, add explicit bullish/bearish flag AnimateDread 2026-07-16 19:55:51 -04:00
  • a4af9313d4 chore(settings): update settingsh1.set configuration AnimateDread 2026-07-16 19:25:26 -04:00
  • fbe4821df2 fix: replace class-balance oversample replay with loss weighting AnimateDread 2026-07-16 19:04:56 -04:00
  • 4f0108366f fix: remove undocumented input noise and unbounded/duplicate features AnimateDread 2026-07-16 18:09:06 -04:00
  • 10f52084a8 fix: training progress Comment() showed the true label, not the prediction AnimateDread 2026-07-16 17:46:27 -04:00
  • 819066819d fix: feature window never included the most recent bars near the prediction target AnimateDread 2026-07-16 15:23:14 -04:00
  • e39cfa56c1 fix: stop halving the WarriorCPU.dll thread pool for phantom concurrency AnimateDread 2026-07-16 13:51:28 -04:00
  • feb82a61aa feat: log per-class OOS recall in the periodic training progress line AnimateDread 2026-07-16 12:25:05 -04:00
  • 78ef4c14b9 fix: classification training was structurally unable to converge AnimateDread 2026-07-16 08:55:54 -04:00
  • 4e3c71ba5f Merge branch 'main' of forge.mql5.io, keeping ADZigZag-indicator labeling AnimateDread 2026-07-16 08:50:51 -04:00
  • eee45526a5 fix: remove redundant Expert.Deinit() in OnInit(); correct ADZigZag buffers AnimateDread 2026-07-16 01:12:37 -04:00
  • 0bfc441139 feat: real-ZigZag training labels, HYBRID CPU-oversubscription fix, input cleanup AnimateDread 2026-07-16 00:56:33 -04:00
  • f115453af2 fix: correct neuron reduction factor logic and add new configuration enums AnimateDread 2026-07-16 00:04:25 -04:00
  • 8c66a0fb89 fix(cl): stabilize network training with tighter limits, weight decay, and sign-agreement gate AnimateDread 2026-07-15 21:47:37 -04:00
  • c05d9d1a1f fix: correct OpenCL gradient and weight update for classification layers AnimateDread 2026-07-15 21:47:09 -04:00
  • ac88fc1ac5 fix(Expert/ExpertSignalAIBase): add feature cache and fix oversampling imbalance AnimateDread 2026-07-14 22:49:14 -04:00
  • 068495b3d7 fix: clean up EA init/deinit lifecycle and close a DirectML mutex-poisoning hang AnimateDread 2026-07-14 22:36:27 -04:00
  • ad154a4897 Merge branch 'main' of https://forge.mql5.io/animatedread/Warrior_EA AnimateDread 2026-07-14 18:05:43 -04:00
  • 1204b9df97 feat: add percentage-based CPU load input for fallback tier AnimateDread 2026-07-14 18:04:48 -04:00
  • 5a3310b5cc AnimateDread 2026-07-13 19:19:42 -04:00
  • 78733aa8a1 AnimateDread 2026-07-13 18:10:22 -04:00
  • aabc3d996b config(inputs): reduce default hidden layers and neuron reduction factor AnimateDread 2026-07-13 17:55:01 -04:00
  • 06f1ffe7c6 Merge branch 'main' of https://forge.mql5.io/animatedread/Warrior_EA AnimateDread 2026-07-13 17:02:07 -04:00
  • 8a1fb8b43a fix: remove native compute DLL embedding and extraction due to MQL5 restriction AnimateDread 2026-07-13 16:58:00 -04:00
  • 9c43740b7b AnimateDread 2026-07-13 16:01:09 -04:00
  • 9464e52784 feat: embed compute DLLs as resources and extend Save/Load for training state AnimateDread 2026-07-13 15:59:35 -04:00
  • f2a30aa9cf fix: avoid TANH gradient vanishing by omitting derivative in output error and add era field to serialization AnimateDread 2026-07-13 08:23:30 -04:00
  • ab69bf82f8 fix: unmerged path in Expert/ExpertCustom.mqh AnimateDread 2026-07-13 07:31:47 -04:00
  • 74c7395127 feat: add max-pooling and convolution OpenCL kernels, clean up barrier and signal code AnimateDread 2026-07-13 03:23:39 -04:00
  • 2f5adb4bec AnimateDread 2026-07-13 03:23:07 -04:00
  • 1073262255 AnimateDread 2026-04-20 22:35:14 -04:00
  • 8157c42314 feat: Enhance README and documentation for Warrior_EA project AnimateDread 2026-04-20 19:28:34 -04:00
  • 0a527b0cf9 convert super.admin 2025-05-30 16:35:54 +02:00
  • 5ad9a96a3a Initial commit MetaQuotes Admin 2025-05-30 14:35:53 +00:00